Computing Box Dimensions from Single Perspective Images in Real Time

Abstract

We present a new method for computing the dimensions of boxes from single perspective projection images in real time. Given a picture of a box, acquired with a camera whose intrinsic parameters are known, the dimensions of the box are computed from the extracted box silhouette and the projection of two parallel laser beams on one of its visible faces. We also present a statistical model for background removal that works with a moving camera, and an efficient voting scheme for identifying approximately collinear segments in the context of a Hough Transform. We demonstrate the proposed approach and algorithms by building a prototype of a scanner for computing box dimensions and using it to automatically compute the dimensions of real boxes. The paper also presents some statistics over measurements obtained with our scanner prototype.